# Mistral vs Grok: API prices compared, dated and sourced

Canonical page: https://bertrande.com/en/vs/mistral-vs-grok/
License: CC-BY-4.0, reuse with attribution and a link.

No value below was produced, estimated or completed by a language model. Every figure was read from a source we name, on the date we give.

The cheapest Mistral model we track, Classifier API model (8B), costs $0.040 per million output tokens and $0.040 per million input tokens. The cheapest Grok model, grok-build-0.1, costs $2.00 per million output tokens and $1.00 per million input tokens. At the entry level, Mistral is 50.0 times cheaper than Grok on output tokens. Prices read on 18 September 2026.

## Mistral models

| Model | Input $/1M | Output $/1M | Read on |
|---|---|---|---|
| Classifier API model (8B) | $0.040 | $0.040 | 2026-09-18 |
| Classifier API model (3B) | $0.100 | $0.100 | 2026-09-18 |
| Ministral 3 (3B) | $0.100 | $0.100 | 2026-09-18 |
| Ministral 3 (8B) | $0.150 | $0.150 | 2026-09-18 |
| Ministral 3 (14B) | $0.200 | $0.200 | 2026-09-18 |
| Voxtral Small | by content type | $0.400 | 2026-09-18 |
| Mistral Small 4 | $0.150 | $0.600 | 2026-09-18 |
| Codestral | $0.300 | $0.900 | 2026-09-18 |
| Mistral Large 3 | $0.500 | $1.50 | 2026-09-18 |
| GLM 5.2 (hosted by Mistral) | $1.40 | $4.40 | 2026-09-18 |
| GLM 5.3 (hosted by Mistral) | $1.40 | $4.40 | 2026-09-17 |
| Mistral Medium 3.5 | $1.50 | $7.50 | 2026-09-18 |

## Grok models

| Model | Input $/1M | Output $/1M | Read on |
|---|---|---|---|
| grok-build-0.1 | $1.00 | $2.00 | 2026-09-18 |
| grok-4.20-0309-non-reasoning | $1.25 | $2.50 | 2026-09-18 |
| grok-4.20-0309-reasoning | $1.25 | $2.50 | 2026-09-18 |
| grok-4.20-multi-agent-0309 | $1.25 | $2.50 | 2026-09-18 |
| grok-4.3 | $1.25 | $2.50 | 2026-09-18 |
| grok-4.5 | $2.00 | $6.00 | 2026-09-18 |
| grok-4.6 | $2.00 | $6.00 | 2026-09-18 |


Method: https://bertrande.com/en/methodology/
Report an error: https://bertrande.com/en/corrections/